Our original hypothesis was that glycolipids can be of value in improving the pathological classification of human gliomas. During the current funding period we obtained considerable supporting information for this hypothesis. Most significantly, we found that a decrease in the proportion of gangliosides of the 1b pathway and the presence of 6'-LM1 correlate with shorter survival times. The group of tumors on which these observations were made is the training set. Here we propose; (1) To test these findings independently in a validation set that we are establishing, and then (2) To evaluate their practical usefulness in a large group-wide study that will focus on creating a panel of markers to improve upon the classification of human gliomas in a clinically beneficial way. In addition, we will examine several other selected glycolipids and their correlation with diagnosis and/or survival using the training set for which glycolipids and consensus diagnoses of the neuropathology panel are available. Those found to correlate with diagnosis and/or survival in the training set will then be studied further using both the validation set and the group-wide set of gliomas. The goal of these studies is to find a pattern of glycolipids that can provide prognostic information beyond that of currently available methods. Such a panel of diagnostic glycolipid markers will provide the basis for improvements in the pathological classification of human gliomas. We will also continue to cooperate with other members of the Glioma Marker network to advance the goals of the RFA by participating in the Coordinating Committee (Yates, Pearl, Mam-ak), neuropathological review sessions (Yates), statistical collaboration with other institutions (Pearl), and maintenance of a group database (Pearl and Mamrak).